In our previous work, we found evidence of large and persistent efficiency differences among healthcare providers - variations in both inputs (utilization) and outputs (risk-adjusted mortality). These studies were largely limited to cross-sectional analysis of the U.S. Medicare population, but we know much less about the dynamic process by which more or less efficient treatments diffuse across providers or regions over time. Nor do we understand the link between provider efficiency - whether costs or outcomes - and the degree of competition and pricing in the under-65 private-pay or Medicaid markets. Finally, how does provider efficiency in the U.S. compare to other countries - does the U.S. over-treat or do other countries ration care? On the question of dynamics, we first ask: given that health care costs have risen over time, was the money spent on efficient (cost-effective) or inefficient treatments? To address this question, we propose to examine specific conditions such as acute myocardial infarction (AMI) and cardiovascular disease in the Medicare population during 1998-2012, but extend the analysis to other disease. To quantify the idea of changes in efficiency, we develop an efficiency index for hospitals and physician-hospital networks and examine changes over time across regions in response to physician and market characteristics. Second, we focus on the interaction among Medicare, Medicaid, and (in Texas) Blue-Cross/Blue-Shield insurance claims, and how differences in prices, market power, and Medicaid payment policies could affect efficiency in their treatment of Medicare patients, whether through effects on quantities (substitution into Medicare when Medicaid pays little) or quality. Third, we compare efficiency between the U.S. and Canada, a country with fee-for-service reimbursement for physicians but global payments to hospitals. We first normalize prices for treatments to abstract from price differences across countries, and compare the distribution of our efficiency index across providers in both countries. Fourth, returning to U.S. data, we develop a new approach to estimating the relative importance of productive inefficiency (differences in hospital expertise holding costs constant) versus allocative inefficiency (treatments with marginal health benefits and poor cost-effectiveness). We use two data sources; (a) detailed chart review data for AMI patients, and (b) surgical quality in Michigan from Subproject 3.